It is well known that chemical carcinogens have the capacity to form covalent DNA adducts directly or indirectly through their metabolites. The formation of these adducts is considered to be the common mechanism by which structurally diverse chemicals ultimately produce mutations and cancer. Identification of DNA adducts might provide a way to monitor human exposure to potential genotoxic agents. We had shown earlier that human lymphocytes incubated with 10-6 M benzo(a)pyrene, an environmental procarcinogen, yield only major adduct, both by high performance liquid chromatography and 32P-postlabeling analysis. Marked variation (less than 10 fold) in the level of adduct formation was observed in lymphocytes from different donors; no significant difference in adduct formation in vitro was observed between smokers and nonsmokers. These findings suggest that lymphocytes can be used to assess the capacity of compounds to react covalently with DNA in human cells. We also detected DNA adducts in lymphocytes analyzed directly by the 32P-postlabeling method that was modified to give enhanced sensitivity by treatment of nucleotides with nuclease P1. Thin layer chromatography maps revealed multiple adducts at variable levels in all specimens, including those from nonsmokers; adduct occurrence was approximately 1 per 107 - 108 nucleotides. One of the adducts co- migrated with the major adduct obtained in benzo(a)pyrene-treated lymphocytes. The results suggest that aromatic hydrocarbon-related adducts accumulate in nonsmokers as well as smokers and that the profile and extent of these adducts can vary greatly among individuals. This study is being extended to include the detection of DNA adducts in human breast tissue. Although estrogens can promote growth through the mammary gland, perhaps through a growth factor pathway, the etiologic factors that make this organ, or other estrogen-responsive organs, susceptible to cancer are not known. The DNA adduct level and pattern of breast tissue obtained by reduction mammoplasty will be correlated with patient profile factors related to known risk factors for breast cancer.